Neural Architecture Search (NAS) is widely used to automatically design the neural network with the best performance among a large number of candidate architectures. To reduce the search time, zero-shot NAS aims at designing training-free proxies that can predict the test performance of a given architecture. However, as shown recently, none of the zero-shot proxies proposed to date can actually work consistently better than a naive proxy, namely, the number of network parameters (#Params). To improve this state of affairs, as the main theoretical contribution, we first reveal how some specific gradient properties across different samples impact the convergence rate and generalization capacity of neural networks. Based on this theoretical analysis, we propose a new zero-shot proxy, ZiCo, the first proxy that works consistently better than #Params. We demonstrate that ZiCo works better than State-Of-The-Art (SOTA) proxies on several popular NAS-Benchmarks (NASBench101, NATSBench-SSS/TSS, TransNASBench-101) for multiple applications (e.g., image classification/reconstruction and pixel-level prediction). Finally, we demonstrate that the optimal architectures found via ZiCo are as competitive as the ones found by one-shot and multi-shot NAS methods, but with much less search time. For example, ZiCo-based NAS can find optimal architectures with 78.1%, 79.4%, and 80.4% test accuracy under inference budgets of 450M, 600M, and 1000M FLOPs on ImageNet within 0.4 GPU days.
相關內容
With well-selected data, homogeneous diffusion inpainting can reconstruct images from sparse data with high quality. While 4K colour images of size 3840 x 2160 can already be inpainted in real time, optimising the known data for applications like image compression remains challenging: Widely used stochastic strategies can take days for a single 4K image. Recently, a first neural approach for this so-called mask optimisation problem offered high speed and good quality for small images. It trains a mask generation network with the help of a neural inpainting surrogate. However, these mask networks can only output masks for the resolution and mask density they were trained for. We solve these problems and enable mask optimisation for high-resolution images through a neuroexplicit coarse-to-fine strategy. Additionally, we improve the training and interpretability of mask networks by including a numerical inpainting solver directly into the network. This allows to generate masks for 4K images in around 0.6 seconds while exceeding the quality of stochastic methods on practically relevant densities. Compared to popular existing approaches, this is an acceleration of up to four orders of magnitude.
The flocking motion control is concerned with managing the possible conflicts between local and team objectives of multi-agent systems. The overall control process guides the agents while monitoring the flock-cohesiveness and localization. The underlying mechanisms may degrade due to overlooking the unmodeled uncertainties associated with the flock dynamics and formation. On another side, the efficiencies of the various control designs rely on how quickly they can adapt to different dynamic situations in real-time. An online model-free policy iteration mechanism is developed here to guide a flock of agents to follow an independent command generator over a time-varying graph topology. The strength of connectivity between any two agents or the graph edge weight is decided using a position adjacency dependent function. An online recursive least squares approach is adopted to tune the guidance strategies without knowing the dynamics of the agents or those of the command generator. It is compared with another reinforcement learning approach from the literature which is based on a value iteration technique. The simulation results of the policy iteration mechanism revealed fast learning and convergence behaviors with less computational effort.
The total generalized variation extends the total variation by incorporating higher-order smoothness. Thus, it can also suffer from similar discretization issues related to isotropy. Inspired by the success of novel discretization schemes of the total variation, there has been recent work to improve the second-order total generalized variation discretization, based on the same design idea. In this work, we propose to extend this to a general discretization scheme based on interpolation filters, for which we prove variational consistency. We then describe how to learn these interpolation filters to optimize the discretization for various imaging applications. We illustrate the performance of the method on a synthetic data set as well as for natural image denoising.
As the methods evolve, inversion is mainly divided into two steps. The first step is Image Embedding, in which an encoder or optimization process embeds images to get the corresponding latent codes. Afterward, the second step aims to refine the inversion and editing results, which we named Result Refinement. Although the second step significantly improves fidelity, perception and editability are almost unchanged, deeply dependent on inverse latent codes attained in the first step. Therefore, a crucial problem is gaining the latent codes with better perception and editability while retaining the reconstruction fidelity. In this work, we first point out that these two characteristics are related to the degree of alignment (or disalignment) of the inverse codes with the synthetic distribution. Then, we propose Latent Space Alignment Inversion Paradigm (LSAP), which consists of evaluation metric and solution for this problem. Specifically, we introduce Normalized Style Space ($\mathcal{S^N}$ space) and $\mathcal{S^N}$ Cosine Distance (SNCD) to measure disalignment of inversion methods. Since our proposed SNCD is differentiable, it can be optimized in both encoder-based and optimization-based embedding methods to conduct a uniform solution. Extensive experiments in various domains demonstrate that SNCD effectively reflects perception and editability, and our alignment paradigm archives the state-of-the-art in both two steps. Code is available on //github.com/caopulan/GANInverter/tree/main/configs/lsap.
Several recently proposed code-based cryptosystems base their security on a slightly generalized version of the classical (syndrome) decoding problem. Namely, in the so-called restricted (syndrome) decoding problem, the error values stem from a restricted set. In this paper, we propose new generic decoders, that are inspired by subset sum solvers and tailored to the new setting. The introduced algorithms take the restricted structure of the error set into account in order to utilize the representation technique efficiently. This leads to a considerable decrease in the security levels of recently published code-based cryptosystems.
Since their introduction the Trasformer architectures emerged as the dominating architectures for both natural language processing and, more recently, computer vision applications. An intrinsic limitation of this family of "fully-attentive" architectures arises from the computation of the dot-product attention, which grows both in memory consumption and number of operations as $O(n^2)$ where $n$ stands for the input sequence length, thus limiting the applications that require modeling very long sequences. Several approaches have been proposed so far in the literature to mitigate this issue, with varying degrees of success. Our idea takes inspiration from the world of lossy data compression (such as the JPEG algorithm) to derive an approximation of the attention module by leveraging the properties of the Discrete Cosine Transform. An extensive section of experiments shows that our method takes up less memory for the same performance, while also drastically reducing inference time. This makes it particularly suitable in real-time contexts on embedded platforms. Moreover, we assume that the results of our research might serve as a starting point for a broader family of deep neural models with reduced memory footprint. The implementation will be made publicly available at //github.com/cscribano/DCT-Former-Public
The generalized inverse Gaussian-Poisson (GIGP) distribution proposed by Sichel in the 1970s has proved to be a flexible fitting tool for diverse frequency data, collectively described using the item production model. In this paper, we identify the limit shape (specified as an incomplete gamma function) of the properly scaled diagrammatic representations of random samples from the GIGP distribution (known as Young diagrams). We also show that fluctuations are asymptotically normal and, moreover, the corresponding empirical random process is approximated via a rescaled Brownian motion in inverted time, with the inhomogeneous time scale determined by the limit shape. Here, the limit is taken as the number of production sources is growing to infinity, coupled with an intrinsic parameter regime ensuring that the mean number of items per source is large. More precisely, for convergence to the limit shape to be valid, this combined growth should be fast enough. In the opposite regime referred to as "chaotic", the empirical random process is approximated by means of an inhomogeneous Poisson process in inverted time. These results are illustrated using both computer simulations and some classic data sets in informetrics.
Residual networks (ResNets) have displayed impressive results in pattern recognition and, recently, have garnered considerable theoretical interest due to a perceived link with neural ordinary differential equations (neural ODEs). This link relies on the convergence of network weights to a smooth function as the number of layers increases. We investigate the properties of weights trained by stochastic gradient descent and their scaling with network depth through detailed numerical experiments. We observe the existence of scaling regimes markedly different from those assumed in neural ODE literature. Depending on certain features of the network architecture, such as the smoothness of the activation function, one may obtain an alternative ODE limit, a stochastic differential equation or neither of these. These findings cast doubts on the validity of the neural ODE model as an adequate asymptotic description of deep ResNets and point to an alternative class of differential equations as a better description of the deep network limit.
Attributed graph clustering is challenging as it requires joint modelling of graph structures and node attributes. Recent progress on graph convolutional networks has proved that graph convolution is effective in combining structural and content information, and several recent methods based on it have achieved promising clustering performance on some real attributed networks. However, there is limited understanding of how graph convolution affects clustering performance and how to properly use it to optimize performance for different graphs. Existing methods essentially use graph convolution of a fixed and low order that only takes into account neighbours within a few hops of each node, which underutilizes node relations and ignores the diversity of graphs. In this paper, we propose an adaptive graph convolution method for attributed graph clustering that exploits high-order graph convolution to capture global cluster structure and adaptively selects the appropriate order for different graphs. We establish the validity of our method by theoretical analysis and extensive experiments on benchmark datasets. Empirical results show that our method compares favourably with state-of-the-art methods.
Deep Convolutional Neural Networks (CNNs) are a special type of Neural Networks, which have shown state-of-the-art results on various competitive benchmarks. The powerful learning ability of deep CNN is largely achieved with the use of multiple non-linear feature extraction stages that can automatically learn hierarchical representation from the data. Availability of a large amount of data and improvements in the hardware processing units have accelerated the research in CNNs and recently very interesting deep CNN architectures are reported. The recent race in deep CNN architectures for achieving high performance on the challenging benchmarks has shown that the innovative architectural ideas, as well as parameter optimization, can improve the CNN performance on various vision-related tasks. In this regard, different ideas in the CNN design have been explored such as use of different activation and loss functions, parameter optimization, regularization, and restructuring of processing units. However, the major improvement in representational capacity is achieved by the restructuring of the processing units. Especially, the idea of using a block as a structural unit instead of a layer is gaining substantial appreciation. This survey thus focuses on the intrinsic taxonomy present in the recently reported CNN architectures and consequently, classifies the recent innovations in CNN architectures into seven different categories. These seven categories are based on spatial exploitation, depth, multi-path, width, feature map exploitation, channel boosting and attention. Additionally, it covers the elementary understanding of the CNN components and sheds light on the current challenges and applications of CNNs.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191